Using encoders to make a 2D measuring device. Where does a newcomer start?

Electronics Arduino
#1

Hello everyone. I am new to the forum and fairly new to electronics. While I have some experience with CNC and 3D printing I m very limited when it comes to the fascinating world of electronics.

Here is the project I am thinking of (and need a lot of help with):

I want to use 2 encoders (one rotational (absolute 360º, one wire actuated, perhaps 1000mm wire length).

The goal is to determine a point in space (actual just 2D) by knowing the angle and the distance from a given point 0. Creating multiple points from the same point 0 should enable me to create a contour.

Easiest example: Rectangle.

First point : 45º 100mm
Second point : 135º 100mm
Third point : 225º 100mm
Fouth point : 315º 100mm

describes a square of roughly 141.42mm

It becomes interesting when the shape describes is irregular and includes curves.
but before I can look into using these measured points I have to measure them first.

The idea is to have a rotating base (absolute rotating encoder) and to mount the wire actuated encoder on top so the wire can be pulled until it reaches the point. Of course the mechanics need to be figured out and I can put my CNC and 3D printing to good use there.

But first I have to understand how to work with these encoders, what kind of output I will get and what to do with them. It seems there are some protocols used but at the moment they are all greek to me. As a former woodworker I am not a mechatronic. At my time (I´m in my 60ies and retired now) that word has not be in existence ;).

I have the feeling this might be an arduino project and I hope someone here can point me in the right direction how that can be pulled of or where something similar already exist.

I appreciate any suggestion

#2

Do you already have encoders, or are you looking for ideas?

Is this enough resolution for azimuth?

#3

Hi Michael thanks for your reply. I don’t have decided on any encoders as I do have to understand first what to look for.

I think in an ideal world a resolution of 0.1° for the rotating encoder would be very nice/ precise. For what I see there is a choice of analog and digital. The analog is to my understanding a kind of potentiometer, the digital reads an internal disc reading out 0 & 1.

The wire actuated encoder is basically as well a rotating encoder with the rotation created by the pull of the wire. For that one I would hope for a 0.1mm resolution…
These specs might be higher than found in the hobbyist range and I might to look rather in the industrial section. Please correct me if I’m wrong.

#4

Well, it’s just math how big a cylinder to use for the wire pay out to get the resolution you desire given the angular resolution of a rotational encoder. That said, I have trouble believing you’ll get repeatable 0.1° azimuth accuracy from a wire alone, so I assume you’ll be running it down an arm and measuring the angle of the arm. And that will give you an azimuth resolution of about 1.75 mm, which contrasts strangely with your desire for 0.1mm resolution on the wire. That’s of by a factor of about 20. And how accurate your azimuth is is a factor of precision, not just resolution. I’d be super surprised if you could repeat an azimuth measurement within 2mm of arc. Would 3.5mm precision on azimuth satisfy you?

It’s easy and reasonably cheap to buy an iGaging style dro up to 600mm, which would probably be a more accurate measure of length and can be interfaced; the protocol is known and I think libraries are available. I don’t see them available at 1000mm though.

With a 1000mm wire you could measure only a 707mm sides (1000/sqrt(2)), and stacking two dros on a Cartesian frame is probably easier. Do you need more than 600mm on a side?

It’s not horribly more expensive for cheap glass 5 micron scales these days, and those easily go beyond 1000mm. Either as the arm of a polar device or as both sides of a Cartesian stack up with linear rails, a glass scale or two would make sense.

Can you describe the motivation for the polar design instead of Cartesian? If I were trying to digitize at that scale and resolution, I would think of ordering linear rail (“chiwin” is probably fine for this application) and glass scales.

1 Like
#5

Michael, thanks again with all your advise. In a different scenario I would agree and measuring with the glass sticks I a carthesian way would make sense. Still I see there still just X and Y addressed.

You are correct I am planing to run the wire inside a telescope tube (at least in the design I have in my head). The tube has a tip and therefore the point to be measure can be clearly defined by angle and distance. The wire alone would not work as the angle of the base has to be exactly lined up with the direction the wire is pulled.

Unfortunately I will need to measure the bottom of a 5-side enclosed space and only have access from above.
Also the space can be as narrow as 6" but also as wide as perhaps 70". Defining point 0 at the center my device need to range from 3" to 35".

As an illustration:
Imagine I want to cut a bottom for an existing box to store and transport hats. Let’s say this is Texas and the hat is a stetson and the box follows the shape of the hat with a certain offset.

The only idea I came up with to take such measurements is my polar device with the telescope measuring arm.

I have seen something similar many years ago and it was used to measure kitchen countertops. Of course a different scale and a different approach without the limitation to put this thing on the bottom of the part to measure…

I my professional life I used big woodworking machines and the precission of the encoders used in these machines were very good. precision of 0.1mm was never an issue.
Today these wire actuated “length” decoders are used in pneumatic and hydraulik applications. I just assume in these automated scenarios things have to be very precise as well.

As I mentioned before I looked left an right and couldn’t find a solution in any price range (not that I could afford any devices like FERO digitisers etc.

The application seems to be very special and there is no desire for the industry to think about it.

Today everything is 3D and massive dataclouds are generated with all kind of scanners. For my application these things are just massive overkill, way too pricey and still often wouldn’t work as they ment to look at the object to scan from the outside not from within…
But I always loved a challenge :wink:

I hope I was able to clear things up a little…

1 Like
#6

Then the rotary encoder I found would have a 0.75mm resolution at 1000mm; it’s about twice the angular resolution at about 0.05°. I don’t know its accuracy. You could use it also for a spring-loaded reel. I’d think you would want to wind Spectra or Dyneema on the reel and choose a way to guide the line not to overlap.